$0.005 per 1M tokens.
There is no free API; the cheapest paid price is $0.005 per 1M input tokens. The weights are open, so you can run it yourself.
#130 of 200 in Embeddings.
On MTEB English v2 · Pair Classification · English, All Mpnet Base v2 scores 83 vs 83.1 — about the same. It is 96% cheaper.
Through the sentence-transformers API.
Released 2021-08-30. It is the current version in its line.
Yes, the weights are open (Apache-2.0 licence).
The weights are open and can be self-hosted; hardware needs are not confirmed yet.
All Mpnet Base v2 is an open-weights embeddings & retrieval model from sentence-transformers, released Aug 2021.
EmbeddingGemma 300M (cheaper), QZhou Embedding (open weights).
CheaperEmbeddingGemma 300MunstatedOpen+5%-60%$0.002Languages: not verified+1KOpenQZhou Embeddingunstated+11%——Memory (4-bit): 0.1 GB → 4.2 GB
BetterBge Base En v1.5unstatedOpen+4%same$0.005iBetterE5 Base v2unstatedOpen+3%same$0.005
Qwen 3 Embedding 8BunstatedOpen+5%+100%$0.01Languages: not verified+1
NV Embed v2unstatedOpen+7%——Memory (4-bit): 0.1 GB → 4.7 GBaLGAI Embedding PreviewunstatedOpen+7%——Memory (4-bit): 0.1 GB → 4.3 GB